2.3 Survey design
Plateau pikas are social animals that usually live in low and open habitats, and their young offspring often stay with the family in their birth year since they are philopatric (Wang et al., 2020; Yu et al., 2017b). Therefore, plateau pikas are territorial and patchy in a vast alpine grassland. Because the diffusion of plateau pikas is a gradual process (Pang et al., 2020b), it is easy to find sites without plateau pika disturbance, even though these sites might be potentially suitable habitats for plateau pikas.
A stratified random and paired design was used to select the plots. At each of the five sites, this study first selected 10 disturbed plots where plateau pikas were present or active burrow entrances were observed. These disturbed plots were 3 to 5 km from each other. Second, a paired adjacent undisturbed plot without the presence of plateau pikas and active burrow entrances was selected for each disturbed plot. The distance between each disturbed plot and its paired undisturbed plot ranged from 500 to 1000 m. If the distance between each disturbed plot and its paired undisturbed plot was too close, plateau pikas could possibly move from the disturbed plot to the undisturbed plot. To ensure that each disturbed plot had a paired undisturbed plot to the greatest extent, each paired plot was ensured to share the same alpine grassland, with no obvious differences in soil types, topography or microclimate. In total, there were 10 pairs of plots at each site and 100 plots across five sites, including 50 disturbed plots and 50 undisturbed plots. The size of each plot was 35 m×35 m, similar to the average area of plateau pika’s home range of 1262.5 m2 (Fan et al., 1999). Owing to grazing from mid-October to early April, the litter in each plot was consumed by livestock, which was better for estimating the aboveground biomass during the warm season.